Re: [PATCH] of/flattree: Use common ALIGN() macro instead of archspecific _ALIGN

[Grant Likely <grant.likely@secretlab.ca>]

On Fri, Jun 25, 2010 at 5:36 PM, Stephen Neuendorffer
<stephen.neuendorffer@xilinx.com> wrote:
>
>
>> -----Original Message-----
>> From: glikely@secretlab.ca [mailto:glikely@secretlab.ca] On Behalf Of Grant Likely
>> Sent: Friday, June 25, 2010 4:34 PM
>> To: Stephen Neuendorffer
>> Cc: benh@kernel.crashing.org; monstr@monstr.eu; microblaze-uclinux@itee.uq.edu.au; devicetree-
>> discuss@lists.ozlabs.org; linux-kernel@vger.kernel.org; jeremy.kerr@canonical.com;
>> sfr@canb.auug.org.au
>> Subject: Re: [PATCH] of/flattree: Use common ALIGN() macro instead of archspecific _ALIGN
>>
>> On Fri, Jun 25, 2010 at 5:18 PM, Stephen Neuendorffer
>> <stephen.neuendorffer@xilinx.com> wrote:
>> >> -----Original Message-----
>> >> From: glikely@secretlab.ca [mailto:glikely@secretlab.ca] On Behalf Of Grant Likely
>> >> Sent: Friday, June 25, 2010 4:16 PM
>> >> To: Stephen Neuendorffer
>> >> Cc: benh@kernel.crashing.org; monstr@monstr.eu; microblaze-uclinux@itee.uq.edu.au; devicetree-
>> >> discuss@lists.ozlabs.org; linux-kernel@vger.kernel.org; jeremy.kerr@canonical.com;
>> >> sfr@canb.auug.org.au
>> >> Subject: Re: [PATCH] of/flattree: Use common ALIGN() macro instead of archspecific _ALIGN
>> >>
>> >> On Fri, Jun 25, 2010 at 4:49 PM, Stephen Neuendorffer
>> >> <stephen.neuendorffer@xilinx.com> wrote:
>> >> > Grant,
>> >> >
>> >> > I need this for the PCI project.  I also have a couple of other
>> >> > generalizations of the drivers/of code
>> >> > to get it to run after boot time in an X86 system...
>> >>
>> >> Send me your patches!  :-)  Get them to me soon and I'll queue them up
>> >> for the next merge window.
>> >>
>> >> g.
>> >
>> > Dangit, I just got the code compiling, and now I find out that X86 doesn't support out_be32()...  :)
>> > You'd think that somewhere somebody might have realized that all the duplicated driver macros that
>> > do that are a bad idea...
>>
>> :-)
>>
>> I think iowrite32be should be cross-arch safe.  in_*, out_* was mostly
>> a powerpc/sparc thing.
>>
>> g.
>
> See.. I told you it was a good idea to have that abstracted.. :)

'course the real problem is not the access macros, but rather the fact
that sometime the xilinx registers are wired up as big-endian, and
sometimes as little-endian.

g.